You are on page 1of 28

CHAPTER TEN The Sociology of Feeling and Emotion: Selected Possibilities

Arlie Russell Hochschild

Feminist criticism of sociology generally settles on one of three observations. Sociologists do not write much about women or all the real varieties of women. Sociologists do not write enough about womens concerns. Or, sociologists do not look at social life with what is traditionally considered a feminine eye. The argument of this chapter essentially fits the last type. In addition, it explores three promising areas for a new sociological theory of feeling and emotion. There i s now no sociological theory of feelings and emotions.1 This is not because the people we study do not take as real the fact that they feel. Nor is it because a persons job, sex, age, ethnic background, or religion is known to be unrelated to how he or she feels in certain situations. It is not, in other words, because the data are not there or are not potentially sociological. Further, it is not because sociologists in their work have completely ignored how actors feel. Ethnography, experimental social psychology, and qualitative sociology generally touch on the concepts of emotion and feeling in the process of explaining why people do what they do and think what they think. What we are missing is some integration of existing knowledge in the form of a theory.


28 1

Perhaps the main reason sociologists have neglected feeling is that, as sociologists, we are members of the same society as the actors we study, and we share their feelings and values. Our society defines being cognitive, intellectual, or rational dimensions of experience as superior to being emotional or sentimental. (Significantly the terms emotional and sentimental have come to connote excessive or degenerate forms of feeling.) Through the prism of our technological and rationalistic culture, we are led to perceive and feel emotions as some irrelevancy or impediment to getting things done. Another reason for sociologists neglect of emotions may be the disciplines attempt to be recognized as a real science and the consequent need to focus on the most objective and measurable features of social life. This coincides with the values of the traditional male culture, to which, i n academia, women are (by exclusion) somewhat less exposed. But if we are to bring sociology closer to social reality, we will do it poorly if we close an eye to feeling. We must open the other eye and theoretically organize what we see.
Images of Actors

Much of social science seems to be based on two images of the social actor, which, like all images of actors, focus attention on certain features of social life. The first image is that of the conscious, cognitive actor, which portrays people as consciously wanting something (e.g., money or status) and consciously calculating the merit of various means toward an end. For example, Erving Goffman (1959) takes US into the world of presentations of selves, and more particularly into the world of rational calculation behind such presentations. It is a world of Everyman as conman, a world of impressions to be managed and manipulated toward the end of a shining self-portrait. Consider Gofhans (1959:4) quote from Waller:
It has been reported by many observers that a girl who is called to the telephone in the dormitories wiU



often allow herself to be called several times, in order to give all the other girls ample opportunity to hear her paged. While Goffman shows us how much more we calculate than we thought, he neglects how much more we feel in socially arranged ways than we thought. We are not shown, for example, how socially induced feelings lead us t o rely on calculation. Such calculation is probably not a constant feature of the consciousness of all social actors. For example, those who experience themselves at the bottom of social hierarchies (women, for example) may be more concerned about looking, smiling, or talking right than are those at the top, whose presentation of self rests peacefully upon reputation and title. Like any other image, this one is not wrong, but it is only partly useful in its choice of what to highlight. It seems to assume that individuals clearly know what they want, and to emphasize the having of the goal (not the doubt or triumph attached to it) and the use of the means (not the g u i l t , apprehension, or glee attached to its use). Those who posit a model of a rational actor generally do not deny that actors feel. However, they imply that little is lost when feelings are ignored or tidily bunched under the t e r m ends and means. The second image, indebted to Sigmund Freud, is that of the unconscious, emotional actor. Here the actor is guided by unconscious motivations, and does or thinks things whose meanings are better understood by the social scientist than by the actor. The actor is said to be driven or prompted by a limited number of instincts, impulses, or needs to achieve, afiiliate, or do any number of things that merely surface as ends or means (Seeley, 1967;Marcuse, 1955). Philip Slater, for example, explores the world of unconscious afect (1964, 1966, 1968). Like a sociological psychiatrist, he focuses on the subterranean channels through which energy emerges into behavior, nearly bypassing the actors consciousness of feeling altogether.* This image, l i e that of the conscious cognitive actor,



does not deny affective consciousness; images deny nothing. Rather the focus on conscious thinkiig, as with Goffman, and the focus on unconscious promptings, as with Slater, allow conscious feeling to fall into a no-mans-land in between. Therefore, we need a third image-that of the sentient actor who is both conscious and feeling. Actors must be seen as more than bloodless calculators or blind expressers of uncontrolled emotions. Human beings, as sentient actors, are aware of their experiences and consciously respond to their feelings and the cultural expectations concerning them. In everyday life, people are often aware of indicating to themselves their subjective states (Imfeeling anxious today) which in turn stand out against a taken-for-granted background stream of experience (My jumpiness is in contrast to my normal calm). Further, the actors select and apply to these states a variety of labels (e.g., anxiety, malaise, uptightness) from among the emotion vocabularies available at their time and in their place in the social world. Every sociological study focuses on a range of variation. The sociology of the sentient actor takes these processes and their products as its focal data. We distinguish between the actors data and their inferences from them, looking to the emotion vocabularies in use, and to background expectancies-whatever the unconscious goings-on beneath the conscious tip of the iceberg. Further, just as certain behaviors (e.g., suicide, homicide, delinquency) are unevenly distributed across the layers of society and the stream of time, so, too, we ask whether and why the various emotions, such as joy or depression, are differentially distributed.3 But are we not caught in a peculiar embarrassment in the elusiveness of our subject matter74 Feelings might seem impossible to capture in the loose net of sociological instruments. Moreover, the actors we study often mistake their feelings. In this case both the experience and the wrong label attached to it are sociological data. If we want to pretend that we know what the actors emotion really is (e.g., Its really depression) and call what the actor t h i n k s it is bias, (I feel sad, energy-less) then



part of our intellectual domain must s t i l l be precisely this bias. For in ridding ourselves of the actors own codification of affective experience, his or her ignorance or linguistic habits, we rid ourselves also of what is social about emotion. We define away from the start what we can then claim we do not find, a sociology of feeling and emotion. We are then left with inferences about instinct or motivation on one side and cognition on the other, because we havent posed our question in a way that would lead to anything else. This image of the sentient actor suggests then that sociologists should attend to the actors own delinition of his or her feelings in order to find out how emotion vocabularies are used, what inner experiences they refer to, and what social situations or rules call them forth or squash them out. The image does not imply that there are no unconscious forces that lead actors to feel as they do. It does not imply that being emotional or emotionless in certain situations is good or bad, sick or healthy. It is often rational, in the larger sense of being adaptive or effective, to feel, just as it could be maladaptive not to feel. M a x Weber misguides us on t h i s point (1966). He confuses irrationality as it refers to behavior, with irrationality as it refers to feeling. He posits a model of social action that is rational, while action based on emotion, like action based on ignorance or tradition, is nonrational. I see two problems with this: a confusion between rationality and emotionlessness, and the implication that emotions and feelings are not positively required by the rational action of individuals and the smooth functioning o f institutions. Weber thought emotions important, and deplored a rationalistic bias that might grow out o f what he meant merely as a methodological device. But I do not get the sense that he saw how necessary emotions were to making things r u n . Take his example of a theoretically posited course of rational action on the stock market. He treats deviations from rational behavior as something the sociologist might explain in terms of irrational emotions (e.g., panic). But the daerence (in terms of feeling and emotion) between the normal stock market and the sudden



depression in stocks is the difference between one affective state of stockbrokers and another affective state. It is highly questionable whether emotion enters into the life of stockbrokers only when there is panic, that emotion makes people act only irrationally. Surely emotion and sentiment are active ingredients in rational behavior as well. A normal day at the stock market, not simply during a panic, would amply show that feelings of excitement, anxiety, or glee are aU part of a good (rational) day's work. Weber mistakes actual emotionlessness for the prevailing norm of affective neutrality we suppose stockbrokers to have internalized. The image of the sentient actor, on the other hand, guides our eye not only to exceptional waves of emotion, as in stock market panics and riotous crowds, it also leads us to notice normal emotions in normal office, factory, school, and home settings.

A Review of Some Relevant Literature

The literature that is relevant to my concern is scattered through various branches of the social sciences.5 It generally reflects one of three approaches to the links among social structure, emotion, and sentience (i.e., conscious feeling). In the first approach (associated with the image of the conscious, cognitive actor), the social context and thinking about emotion are linked, but conscious feeling tends to be neglected. In the second approach (associated with the image of the unconscious actor), unconscious emotional phenomena and social structure are liked, but again conscious feeling is omitted. In the third approach, the relation between sentience and its labels is analyzed, but the social context is often neglected.'j The first approach to the sociology of emotion is to study what and how people think about emotion and feeling. This is the concern of attribution theorists who study actors' ideas about the causes of behavior (Jones et a l . , 1972). Some experimental psychologists have studied emotions as actors use them in their attributions of causality. An anthropologist, Robert Levy (1973), also exemplifies this focus. He notes that among Tahitians the emotional re-



sponse to the loss of someone dear is attributed to illness and its nonemotional symptoms. (Romantic love and extreme jealousy are attributed to being crazy and considered somewhat abnormal and bad.) In a somewhat related way Blum and McHugh (1971) deal w i t h motives and not emotions. They dont see these motives as concrete private and interior mainsprings but as a way of conceiving of social action. To the radical ethnomethodologist, the way of conceiving of social action is no small matter, since it determines and constitutes social action. Even if we accept this view of social action, we need to know the actors view of this concrete psychological state of &airs-sentience-in order to know upon what the actor is basing his or her explanations for events. For example, in the study of sex roles, it would be interesting to explore sex differences in motive report. I quit graduate school because I fell i n love with your father is perhaps a more common and legitimate motive report for women than is its counterpart for men. The second approach, corresponding to the second image of the actor, mainly involves various ways of applying Freud to social science (Dollard, et al., 1939; Seeley, 1967; Slater, 1964; Gorer, 1964; Mead, 1949; Erikson, 1950; Malinowski, 1927). While these studies provide an enriching integration of what are fairly differentiated fields (Freudian theory and sociology, say), they still neglect the conscious experience of sentient actors. This results in studies with a simultaneous focus on the unconscious and the social, with conscious feeling edged out now by two sides rather than one. Dollard, for example, uses the emotion word frustration to refer to observable behavior resulting from situations in which expected acts are prevented from occurring ( 1939 :7). Between inducing situation and consequent behavior, Dollard gives only a casual glance at conscious feeling. The it of frustration that gets displaced from one issue to another, the it that is socially caused and in turn causes behavior, remains mysteriously out of view. We learn more from Dollard on the situational side than on the response side. His research nonetheless suggests for the study of sex roles how aggression, once



its experiential referent is speciiied, is displaced in different ways for men and women-the central concern of my third hypothesis (Feeling and the Politics of Aim). The third approach is to spell out in experiential terms what emotion words refer to. Here the inner world of feeling is mapped against the cultural world of labels. This is done either by holding experience constant and examining variations in labeling, or by holding labels constant and examining variation in experience, or by looking at the interrelationship, holding neither constant.7 An example of the last approach is Levys (1973) study of what he called transschematic experience and cultural schema among Tahitians. In a list of 301 words describing feeling in the missionary dictionary, 47 referred to angry feelings and 27 to pleasurable states. To the Western eye, some feelings (e.g., anger, shame, fear) were well discriminated, while others (e.g., loneliness, depression, guilt) were poorly discriminated. In the study of sex difEerences we might determine whether and how the same labels refer to different experiences for men and women. For example, Kephart found that college women reported more infatuations than did college men. He reasoned that for women looking back, love affairs are related to infatuations , and are remembered merely as passing fancies (Kephart, 1967: 472). Because women are the upholders of the monogamous ideal, he reasoned, they tend to relabel or reremember past love as mere infatuation. Associated, then, with the first image of the conscious cognitive actor and the second image of the unconscious emotional actor, we have two lines of inquiry that tend to bypass conscious feeling. Associated with the image of the sentient actor, we have a line of inquiry that seeks to relate feelings to the label world. I think we need to start with the latter and elaborate it in a depth-psychology direction (drawing from Dollard) and in a cognitive direction (drawing from Blum and McHugh). Feeling, thus conceptualized, can then be elaborated in a sociological direction. In doing this, sociologists have focused on the social causes and consequences of a great variety of emotions and senti-

. .



ments.8 On the causal side, some, like Moller (1958) and Lewis (1959) in their studies of love, deal with the grand structural picture, while others, like Gross and Stone (1964) in their study of embarrassment, deal with the immediate interactional settiog. On the consequence side, some, like Foster i n his study of envy (1972), analyze the social forms (customs and institutions) that function to avert envy by devaluing, hiding, or symbolically sharing the envied object. Others deal with the nonstructural, temporary consequences of emotion, such as episodes of mass violence (Gurr and Ruttenberg, 1967)
. Q

Toward a Sociological Theory of Emotion and Feeling Feelings take on their meaning ooly in relation to a specific sociohistorical context. This context can be understood first as a place in the world, a sphere-such as the family or secretarial pool. Feelings in each sphere also take on their meaning with reference to various dimensions of the context. I will explore three here: the normative, the expressive, and the political. The normative dimension has to do with the relation between feeling and feeling rules (this situation makes me happy, but in this situation I shouldnt be this happy). Both feelings and feeling rules are socially induced, and so, consequently, is the potential conflict between the two. The expressive dimension has to do with the relation between a persons feelings and other peoples understanding of those feelings. Here were dealing not with the wrongness of feelings but with the inferred falseness of them. The political dimension concerns the relation between a persons feelings and the target of those feelings. It focuses on the aiming of affect at those higher or lower (more or less powerful) than oneself. Thus, the first dimension tells us about judgments on feeling, the second about cornmunication of feeling, and the third about the aiming of feeling. These three levels of feeling-in-context are suggested by the image of the sentient actor and to some extent by the other two images as well. I . Feeling and feeling rules. Social arrangements in-



duce a range of feelings. They also, in various ways, control these feelings.lO One means of control is via feeling rules, which define what we should feel in various circumstances. Such rules put a normative floor and ceiling on feeling by indicating what is appropriate and desirable. We can posit certain hypothetical ways in which social arrangements set up feelings and feeling rules, and in which the two coincide. We can measure these against our findings and identify deviant patterns that need further explanation. For example,ll a professor, by virtue of the role, is placed in certain feeling-inducing situations, such as receiving many s o r t s of requests from a variety of personalities. The general feeling rule is to care in a mild, delimited way about the students intellectual development. One must not feel angry or hate a student, no matter how obnoxious. The professor who actually hates a student experiences the pinch between feeling and rule. Like the psychiatrist who falls in love with a patient while trained to affective neutrality, the professor is likely to experience such anger or hatred as a violation of the feeling rule. Moreover, women professors must combine the potentially conflicting feeling rules of affective neutrality associated with professionalism, and caring and nurturing associated with femininity. If she feels anger toward a student, both rules pinch experience (WiMer, 1974). The distinction between feeling and rule is less apparent, although equally present, when we feel as we should. We realize the silent presence of feeling rules when we experience good role performances, such as by an actress, receptionist, professor, and other occupations perhaps mainly in the service sector. Their success is based on using feeling rules effectively to induce certain feelings in the audience. Different occupations trade on different types of emotional skill-inducing and expressing emotion for the actress, understanding or interpreting it for the psychiatrist, disguising it for the poker player. Dserent affective areas are thus professionalized depending on the task-the area of trust for the Presidents press secretary, of fear for the



claims agent or jailer, and the squelching or fostering of hope for the college counselor or nurse, Even before behavior occurs, people perform emotion work upon their feelings.12 It is not only behavior that feeling rules govern, but, in varying degrees, feelings themselves. We not only like but we also try to like people we feel we should or must like. Often we not only feel sad at a funeral but we also try to feel sad there, or happy at a marriage. Parallel to the expression work that actors perform to maintain the appearance of self and the social occasion (Goffman,1959) is the deeper work on feeling. In a sense, society induces the self to induce and control feeling such that one continually socializes his or her feeling. Feeling rules channel the stream of subjective experience as it enters different spheres of the symbolic world. Some spheres are generally shared, some are partially shared, and others are private (e.g., the symbol systems of couples). In the most private sphere, some subjective experiences do not seem to be social at all, and the rules have little social bearing. But as the symbolic world changes with structural change, so does the bearing of feeling rules. For example, we may more often think we should feel moved and not be in the abandoned ritual spheres of political or religious life (What a silly ceremony. Why am I so unmoved?). Also, we may more often feel moved in private spheres in ways that, according to the rules, are inappropriate. For example, the other day I saw a TV advertisement for spray antiseptic. For some obscure reason, the picture of the mother ministering to her childs scraped knee moved me. Also part of my reaction, however, was the attitude that TV commercials are mundane and manip ulative. I did not approve of my feeling. It did not seem to me false (the ad worked), only absurdly, inappropriate, and a moment later I dismissed it as silly. There seem to be three types of appropriateness of feeling: (a) Clinical appropriateness refers to what is expectable for normal, healthy actors (e.g., the actor may think her/his fury healthy, despite its moral inappropriateness). (b) Moral appropriateness refers to what is morally legit-


29 1

imate (e.g., the actor may get furious at a helpless child, but this may be morally inappropriate). (c) Social-situationd appropriateness refers to what is called for by the norms specific to the situation (e.g., to feel effervescent at a party). These three types correspond to the actors roles in everyday life as, respectively, a folk clinician, minister, and etiquette expert. In the case of the TV ad, I did not invoke a clinical should (What is wrong with me? What a sick reaction). Nor did the feeling hit upon a social-situational should-in a sense it was all too socially appropriate. It was more my moral evaluation of the intention behind the ad that detached the feeling from legitimacy and made the moment iindy siUy. The sociology of silly moments tells us about the symbolic world in which we share not only rules that we should feel, but also rules that we should not, for some reason, feel what we do feel. Let me briefly apply these ideas to envy and suggest some implications for women. Envy is to competition as jealousy is to hierarchy (Apthorpe, 1972: 187; Simmel, 1950). Competition (the system) induces rivalry (the relationship), which fosters envy (the feeling). While the moral injunction against envy applies to winners and losers alike, envy is unequally distributed among winners and losers. In other words, the socially induced feeling and the rule against it are systematically discrepant, and the discrepancy itself varies from one group or stratum to another. For example, the upper-middle-class, college educated housewife is in a situation likely to induce envy of her husband-say, a corporate executive-who is winning what she was possibly aspire to herself. How envious she feels depends, of course, on a number of things-a socialization that leads her to devalue housework and child rearing and that fosters the revolutionary act of comparing herself to her husband and to men in general. It depends also on her attitude toward the sexcaste arrangements (originally designed, according to Foster, to encapsulate potentially envious populations). Presuming a modern socialization, what inequities persist in



her life wilI attach to envy-a feeling to which low selfesteem makes her especially vulnerable. Envy without a social movement is a particularly private, unlegitimated feeling, attached as it is with the rule Dont. The stream of subjective experience knocks against the private, symbolic world that upholds the couples solidarity. The feeling rule upholds that symbolic world. It privatizes envy which, no matter how powerful, is rendered petty or silly. The discrepancy between socially induced feelings of envy and envy rules may result in a number of social cust o m and institutional arrangements that handle this discrepancy. Foster (1965) suggests that the experience of envy is a result of (a) notions about the limited or unlimited supply of desired goods (money, love, honor, security), (b) their distribution, and (c) the principle of equivalence (the tendency to equalize goods). He suggests several social devices that deal with envy-concealing or denying possession and truly or symbolically sharing it. The housewife in my example might mentally reduce her envy (dwelling on the misfortunes of working women or the hazards of the male rat race). More familiar is the device of symbolic redistribution, such as the custom, common in the upper levels of social life, of the wife appearing at her husbands occupational ceremonies and symbolically sharing his glory. 2. Feeling, expression, and the culture of translation. Just as feelings are linked to rules in a normative context, so feelings are linked to expressions in an expression context. Just as actors make judgments on experience in a context of rules, so too they make judgments upon expression in an expression context. In mapping rule to feeling, the actor judges whether a feeling is appropriate in the clinical, moral, or situational sense. In mapping expression to feeling, the actor judges whether the expression is true or false (i.e., whether it corresponds to a real subjective experience). In both cases, the context itself varies from one category of people to another. When I smile at you, I offer a sign or symbol of my inner feeling: say, liking. When you see me smile, you make a translation



between my outer smile and my inner fee1ing:You must, in a flash, make an inference, correct or not (Does she really like me or is she just being polite? Is this expression true and serious, or should it be discounted?). Quite apart from the judgment about the appropriateness of my liking or yours, there is the task of inferring from my smile to my liking you, and the results of that process also vary with social context. The many small decisions that lead us to discount or take seriously an expression rest on a variety of factors: our style of interpreting, our knowledge of anothers smiling (as opposed to, say, hugging) habits, our knowledge of events prior to the encounter, and so on. These elements also operate within a larger social context i o which some expressions are by custom scarce and others abundant. The general market o f expressions thus influences the value we impute to a particular smile as well as the probability of s true (corresponding to inner feeling) or perceiving it a false. The more intimate the relation, the more that private knowledge affects interpretation. The more a person is a stranger to us, the more influence does public knowledge of the conditions of the general expression market have on our translations. Expressions can be seen as a medium of exchange. The translation between expression and experience can be seen as analogous to the translation between a paper dollar bill, a symbol, and its value (Parsons, 1968). Like paper money, expressions such as a smile are symbolic and in circulation. They are symbolic with reference to certain takenfor-granted agreements as to meaning and institutiona1 arrangements that support this consensus. Like money, expressions work on a basis of trust that this expression (e.g., a clenched fist) corresponds to that range of inner experience (e.g., anger). The expressive context rests on a public trust i n the validity of the translation. There can be a breakdown of confidence in the dollar, as i n a recession or depression. Similarly, there can be analogous recessions or depressions in the value of, or trust in, emotive expression: a shaken confidence in the meaning of a smile or a kiss. ln a bureaucratized, rationalized



society with a large service sector, displays of feeling are also rationalized. Expressions of positive effect-the have a nice day buttons, the waiters hope you enjoy your meal, the receptionists smile-are so abundant that trust in them is reduced. Such expressions are like junk mailings, meaningful (and expected to be profitable) for only a few people and thrown away in a mental wastebasket by most others as false outer expressions corresponding to little inner feeling. I n a commercialized society, positive expression is more inflated than expression, say, of envy, anger, or resentment.18 There are more phony dollars in circulation. A slight expression of anger is trusted to correspond to felt anger in a way not generally true for an expression of liking. Expressions of anger are more serious and more likely to be sensed as true.14 Within the general market of expressions. there are particular markets associated with regional subpopulations or strata. Within the expression market for anger in southern Italy, for example, anger is cheaper than it is among Maine Yankees.15 Moreover, sex-role socialization may render anger expressions more scarce and serious for women than for men. Bugental et al. (1971) show that women are more likely to smile even while angered or frustrated than are men. Eckman (personal communication) in his study of facial expression found women more likely than men to mask (quickly cover up) anger, while men more often mask fear. A study of affect display in childrens stones might show girls to be portrayed as more expressive o f fear than of anger, while for boys the opposite is probably true. In any case, the translation from outer expression to inner feeling and vice versa is set against different expectations about anger expression in women than it is for men. 3. Feeling and the politics of aim. Feelings are linked not only to rules (in normative contexts) and to expressions (in expression contexts), but also to sanctions (in political contexts). On this third level, there is a relation between the distribution of powerle and sanction on one hand and the target of feeling and expression in social sit-



uations. Whereas the first two levels deal with conscious feeling and thought (the sentient and cognitive actor), this third level deals with what is often, although not necessarily, unconscious feeling. The relation between sanction and feeling varies for different feelings. Let us consider anger. Under certain social conditions, anger, insofar as it is deflected at all from its rightful target, tends to be deflected down into relative power vacuums. Thus anger is most likely to be aimed at people with less power, and least likely to be aimed at people with more power. Unlike envy, anger runs in channels of least resistance (Dollard et al., 1 9 3 9 : 4 0 4 ; Levy, 1973:286). The pattern is clearest in the case of the expression of anger. But I think in a milder way it is there also for the very experience of anger. This general pattern parallels the hierarchy of joking observed by Rose Coser. Analyzing conversations in which humor occurred at staff meetings of a mental hospital for three months, Coser concluded, Those who were of higher status positions more frequently took the initiative to use humor. More siflcant, still, the target of a witticism, if he was present, was never in a higher authority position than the initiator (1960:95). Insofar as jokes with a butt are a benign cover for hostility, they reflect the pattern in this way. Contrariwise, more positive feelings tend to run up the sociopolitical hierarchy (e.g., the phenomenon of kissing the hand of the Pope, or honoring the Queen). Under the governance of socially organized fear, there is both the downward tendency of negative feelings and the upward tendency of positive ones.17 I have suggested that when deflected, anger and resentment are normally deflected down. This pattern crosscuts that of the deflection of hostility, suppressed in the in-group and expressed toward the out-group. Insofar as social incentives induce a great deal of resentment and anger (as in times of economic and social insecurity), more resentment will be deflected, and deflected down. Needless to say, not all anger is displaced. Also, for the powerful, doormen and secretaries for instance may



provide a human barrier against exposure to hostility that powerless people less often enjoy. But, even without these barriers to exposure, when affect is displaced it is generally bestowed i n distinctly diferent ways. One reason people want power, and even more so, honor and glory, is p r e cisely t h i s sort of immunity from hostility and exposure to awe and liking. There are important conditions18 under which the "aim" of bestowers is deflected up-for example, in cases of "rebellion." Then the latent, rule-constrained envy of the subordinate attacks the institutional arrangements to avert it. But the emotional equivalent of "false consciousness" (such as feeling content with one's fate) is more the rule than the exception. This up- and downstream pattern has consequences for the affective worlds people inhabit. Those near the bottom of power hierarchies tend to bear a disproportionate amount of displaced anger. Women, for example, receive not only their husband's frustration displaced from the office to home, but also the anger of other women who are similarly displaced upon. If a woman takes her anger down (to children) and occasionally across (to other women), she, by the same token, becomes the less powerful target of both men's and women's anger. The least powerful become the targets of a wide variety of hostility. In a sense, they become the complaint clerks of society, and make a similar art of appeasement. For those on the bottom rungs of the political ladder, the world is experienced more often as a hostile place. Contrariwise, powerful people not only get a disproportionate amount of other resources such as money and prestige, but also enjoy more affective rewards. For the. dwellers at the top, the world is more often experienced as a benign place. As a consequence, powerful and powerless people live in different emotional as well as social and physical worlds. This is perhaps especially so for the extreme cases, where the actor is at the very top or bottom of several (e.g., class, sex, race) stratification systems. This has implications for the study of sex roles. We might, for example, seriously examine the proverbial case



of the boss who blows up at the worker, the worker who blows up at his wife, the wife who gets angry at her children, and the children who take it out on the dog. The very first task would be to explore the conscious feelings of a constellation of actors. The second task would be to determine what anger seems to be displaced and what n o t . We might ask who gets how angry, at whom, for what. When a slightly burned chicken draws a raging response from the husband, and when a childs small miscalculation about continence draws a storm of reaction from the mother, we might develop an educated suspicion about displacement. The third task would be to set the pattern of feelings and inferences about displacement against the conscious background expectations and rules of both the actors and the social scientist. As washtub serials on TV suggest, displacement has in part become normalized as something wives expect from their husbands when there is trouble at work. But what happens to womens anger remains, not only for Freud but also for social science, a mystery. The Paradox of Womens Anger

The fascinating sociological problem, as Daniel Bell notes, is why in democratic society, as inequality decreases, resentment increases (1973:451) .I9 It is by now commonplace to observe the same paradox among women. The upper-middle-class, college-educated woman is more likely to resent the inequality of the sexes than the lowerclass, housebound mother of seven. It was also in the late 1960s, when things were getting somewhat better, rather than in the 195Os, when they were getting somewhat worse, that feminism again began to spread in urban middle-class circles. The usual explanation for this paradox is the revolution of rising expectations. Against newly increased background expectations of what is just and normal, the remaining inequities glare even more sharply. This is a good, but partial and strangely cognitive, explanation. It leaves us to wonder just how anger or resentment attaches



itself to these glaring inequalities. Attached to roughly the same perception of inequality and expectations are a variety of emotional states: Some women appear to raise their expectations without seeming to experience anger or resentment; some experience overt anger aimed at the oppressive target, but others appear to turn blame on themselves or on other women; and some experience only a vague discontent that dwells on nothing in particular but everything in general. The variation is not entirely explained by differences in personal biography. Even women with similar personalities may well feel differently about perceived inequality, and the same woman may, needless to say, channel her anger a e r e n t l y at various times. At least part of the paradox might be explained by development of the feminist movement itself. The feminist movement appeals to the relatively deprived, who have some sense of their rights, but experience a frustration in claiming them. They less commonly attract the haves and, at least not at first, the have-nots. Once under way, such a movement reverses the pattern of deflecting anger down or over (e.g., toward other women) by a unionization at the middle and bottom and by publicizing ideas that foster a simultaneous identification with the rights of men and d e identification with men themselves. (I want those rights, but 1 dont have to be a man or be l i e a man to deserve them.) Movement ideology touches different psychic bases and different crystallizations of power and sanction in family and workplace. It touches different sources of readiness to feel anger and act among different strata of women. For some it channels existing anger, for others it brings out latent anger, and for still others it legitimates old feelings with new feeling rules. In widening circles, anger becomes more legitimate and intelligible. It is even positively required for full membership, for there is now something false about neutral or positive feelings. Social movements for change make bad feelings okay, and they make them useful. Depending on ones point of view, they make bad feelings rational. They also make them visible. It is common to associate emotionality with social move-



ments, and not with the ordinary goings-on in schools, hospitals, factories, and corporations. A sociological study of emotion and feeling would remind us that emotion is as salient an issue when it appears in the office as in the street, as much a part of everyday experience for the Establishment as for the angry young woman. Emotion is not the rebels specialty, not an exceptional wave in an essentially calm sea. The rebel simply works by different rules of feeling, different standards of truth and falseness of expression, and exemplifies different patterns of aim. In not feeling as she should, the rebel reveals what emotional ways we conventionally take for granted. Just as occasional traffic accidents make us marvel at the normal absence of them, so the emotional rebel who does not feel as she should makes us marvel that rebellion is so rare.

I have suggested in this chapter two underlying images in much of sociology: that of the conscious, cognitive actor and that of the unconscious, emotional actor. Between these two lies the missing image of the sentient actor, to which I tie the possibility for a sociological study of emotion and feeling. I have suggested that our data constitute conscious feeling, reflected in introspective report. This report involves both the data of experience against a backdrop of what passes as normal experience. It involves also the inferences from data to which people attach labels from the culturebound vocabularies of feeling. We base ow ideal typical paths of social arrangement, feeling, and consequence on the concept of emotion and its function for perception, cognition, and behavior. We might trace such patterns as they bear on the rules we apply to our feeling (in rule contexts), the judgments we make about other peoples feelings (in expression contexts), and the ways in which we direct or misdirect feeling (in sociopolitical context). I suspect that the social goes deeper than our current images of the social actor have led us to suppose. Roles and relations are surely not social patterns that apply only



to thought and action, leaving feeling an untouched,

timeless, and universal constant. We must integrate the sociology of the head with a sociology of the heart and somehow erase the distinction in the process.

I would like to gratefully acknowledge Jessie Bernard, Wally, Adam Hochschild, Gertrude Jaeger, Alice Rossi, Ann Swidler, and Norma Wikler for helpful comments on an earlier draft of this chapter.

1. I d e k e emotion as conscious awareness of bodily co-operation with an idea, thought, or attitude and the label attached to that awareness. 2. For example, Slater (1968) suggests the Greek captive-slave mother takes out unconscious aggression on her son, creating homosexuality in him. 3. As Allport and Odbert observe, certain terms came into use in English only after the eighteenth century (e.g., depression, ennui, chagrin, apathy), and the modern sense of some older terms has become more subjective (e.g., constraint, embarrassment, disappointment). Labels are not univocal symbols corresponding through the ages to fixed varieties of human disposition (1936:3). f we focus only on overt behavior, then aspects of 4. I social reality suggested by all three images would be missed, particularly the disjunction between feeling and behavior. For example, Kephart (1967) asked college students, If a boy o r girl had all the other qualities you desire, would you marry this person if you were not in love with himher? A total of 64 per cent of the men, but 24 per cent of the women, said No. For men and women, feeling in love appears to have a different tie to the act of marrying. 5. Some writers find roots in phenomenology (Davitz, 1969; Block, 1957; Sartre, 1948), Freud (Seeley, 1967; Slater, 1964), or Simmel (Klatsky and Teitler, 1973), while others bypass social science altogether and go back to Descartes (Davis, 1936)-a sign surely of an intellectual infancy with a quiet disagreement about parentage. 6. Moreover, each uses emotion differently. Tn the h t approach, it refers to a concept actors use to make



sense of their experience. In the second, emotion refers to a concept the social scientist uses to make sense of the actors experience by referring to the actors unconscious. In the third, emotion also refers to a concept the social scientist uses, but here it is defined as the association between bodily or psychic experience and the cultural meaning and label assigned to it. My preference is to start with the third conceptualization and extend it to incorporate insights drawn from the other two approaches. 7. Schachter exemplifies the first., Davitz the second, and Levy the third approach. For Schachter, emotion is basically physiological arousal with a label attached (Schachter and Wheeler, 1962; Schachter, 1964; Walster, 1971). In one experiment, he gave injections of epinephrine to some subjects and placebos to others. Some subjects were informed, others misinformed, and others told nothing about what physiological experience to expect. Then all subjects were placed in either an anger-inducing situation (in a room with an angry person) or a euphoria-inducing situat i o n (in a room with a man shooting paper planes, etc.). Those who were physiologically aroused by the drug, had no explanation provided, and were exposed t o angry or euphoric company tended to label their own state as like that of their partner. Thus, subjects labeled d8erently the same physiological experience depending on their explanation for it and on the social milieu. Davitz (1969) does the opposite of Schachter. He holds the labels constant and examines variations in the reported experience attached to them. He had respondents complete a list of items for fifty emotions, reporting physical sensations, perceptions of situation, and expressive behavior. For his limited college sample, he found considerable consensus about the link between emotion label and experience, although more so for some emotions. 8. Some deal with jealousy (Davis, 1936), envy (Foster, 1972; Schoeck, 1966), embarrassment (Modigliani, 1968; Gross and Stone, 1964), trust (Deutsch, 1958; Klatsky and Teitler, 1973), aggression and hostility (Berkowitz, 1962; Gun: and Ruttenberg, 1967; Walters, 1966), grief (Averill, 1968), and love (Goode, 1964; Huizinga, 1970; Lewis, 1959; Moller, 1958; Rubin, 1973). 9. Davis classic study of jealousy suggests a kind of analysis revelvant to sex roles. Rejecting the position (of



the family historian, Westermarck) that adultery %aturally arouses jealousy, which in turn causes monogamy, Davis suggests that the cause of jealousy lies in the expectations set up by the institution of monogamy, which then make adultery arouse jealousy. He focuses on the jealousy of men over their wives, who are conceived of as property. This property can be borrowed or lent without jealousy, as occurs in some traditional societies, but when the property is stolen or trespassed upon, jealousy is aroused. This feeling can be seen, then, as a legitimated and appropriate response in light of the prescribed social ties of monogamy and the associated feeling rules that cause the arousal and aiming of jealousy at a particular target. One could also analyze the social consequences of the feeling and feeling rules of jealousy, such as the guarding of property, as well as the effects on these feelings of major social changes, such as sexual equality. 10. These rules are internalized to a greater or lesser extent, are more or less conscious, and correspond to various types of sanctions. Feeling rules must be distinguished from expression rules (Gohan, 1959). For example, it is not sanctioned for a man to feel no fear in a fear-inducing situation. Rather, he should feel fear, hold it in, and appear stoical. Of course, rules are not the only m e a n s of control of feeling. 11. Although I am focusing here on feeling rules associated with a variety of roles, such rules bear on feeling in different ways for different personalities. Such rules also interact with other more general feeling rules associated with sex, religion, and social class. 12. Organizations also do emotion work. Many aspects of bureaucracy (e.g., written communication, secrecy patterns) serve a variety of emotional functions that supposedly foster the smooth flow of work and affective neutrality. 13. Also anger is more disruptive than liking, and this may be one reason why, in a rationalized society, s m i l e s proliferate. Where there is little basis for genuine personal liking, as in many commercial relations, inauthenticity performs a positive social function. I am indebted to Gertrude Jaeger for this point. 14. False has at least two meanings: the discrepancy between perceived display and inferred feeling (artificiality) and the discrepancy between the true or original na-



ture of something and its presenting version (inauthenticity). I use the term in its fist sense. 15. Expression contexts also vary over time, with changes of style in emotional expression. Perhaps we can get at this by examining the film industry, especially bad films and what is popular at different times. Hollywood is to emotional life what the State Department is to foreign relations. Hollywood exercises a certain cultural hegemony over the world of emotion, teaching us how kissing and fighting are done, how feelings are to be expressed and managed. 16. Some treat emotion as the great equalizer, assuming that the powerful do not necessarily enjoy more good feelings than the powerless. I am skeptical, for I doubt that power and emotion are unconnected. 17. In the case of positive emotions going up, there are often latent secondary gains. When a lower-status person positively identifies with a higher-status one, there may be a magical transfer of goods (I have by identification with my boss what I may not have otherwise-her/his power and prestige). See Hochschild, 1973: Chap. 5. 18. Four main factors seem to account for a change in aim: a change in the strength of the incentives for aiming in a given direction, a change in the strength of the controls constraining the aim, a change in the conditions that serve as outlets or safety valves, and a change in extent of contact and affective integration of the actors. 19. We must question whether resentment does, in fact, increase or whether it simply becomes more visible and expressed. Bell also concludes that scarcity is relative and inevitable, and that it thus makes little sense to wonder about institutional arrangements that would alter inequality. I do not see anger and envy as that divorced from structural arrangements, and for this and other reasons, I very much disagree with his conclusions.

Allport, G., and Odbert, H. S . 1936. Trait Names: A Psycho-lexical Study, p ~ y chological Monographs 47: 1-171.



Apthorpe, R. J. 1972. Comments (On Foster, 1972), Current Anthropology 13:2034. AverilI, J. 1968. Grief, Psychological Bulletin 70:72148. Bell, D. 1973. The Coming of Post-Industrial Society. New York: Basic Books. Berkowitz, L. 1962. Aggression. New York: McGraw-Hill. Block, J. 1957. Studies in the Phenomenology of Emotions, Journal of Abnormal and Social Psychology 54:358-63. Blum, A. F., and McHugh, P. 1971. The Social Ascription of Motives, Ainericun Sociological Review 36:98-109. Bugental, D.; Love, L.; and Gianetto, R. 1971. Perfidious Feminine Faces, Jortrnal of Personality and Social Psychology 17:314-18. Coser, R. L. 1960. Laughter Among Colleagues, Psychiatry 23: 81-95. Davis, K. 1936. Jealousy and Sexual Property, Social Forces 14:395410. Davitz, J. 1969. The Language of Emotion. New York: Academic Press. Deutsch, M. 1958. Trust and Suspicion, Journal of Conflict Resolution 2:265-79. Dollard, I., et al. 1939. Frustration and Aggression. New Haven, COM.: Yale University Press. Erikson, E. 1950. Childhood and Society. New York: W . W. Norton. Foster, G. M. 1972. The Anatomy of Envy, Current Anthropology 13:165-202. Goffman, E. 1959. Presentation of Self in Everyday Life. Garden City, N.Y.: Doubleday.



Goode, W. 1964. The Theoretical Importance of Love, The Family, ed. R. Coser. New York: St. Martins Press. Gorer, G. 1964. The American People. New York: W. W. Norton. Gross, F., and Stone, G. 1964. Embarrassment and the Analysis of Role Requirements, American Journal of Sociology 80:1-15. Gurr, T., and Ruttenberg, C. 1967. The Conditions of Civil Violence. Princeton, N.J.: Center of International Studies, Princeton University. Hochschild, A. R. 1973. The Unexpecfed Community. Englewood Cliffs, N.J. : Prentice-Hall. Huizinga, J. 1970. The Waning of the Middle Ages. London: Arnold. Jones, E. E., et al. 1972. Attribution. Morristown, N.J.: General Learning Press. Kephart, W. 1967. Some Correlates of Romantic Love, Journal of Marriage and the Family 29:470-74. Klatzky, S., and Teitler, M. 1973. On Trust, Working Paper 73-79, presented at the annual meeting of the American Sociological Association, August 1973. Levy, R. I. 1973. Tahitians. Chicago: University of Chicago Press. Lewis, C .S . 1959. The Allegory of Love. New York: Oxford University Press. Malinowshi, B. 1927. Sex and Repression in Savage Society. New York: Harcourt, Brace. Marcuse, H. 1955. Eros and Civilization. New York: Vintage. Mead, M. 1949. Male and Female. New York: Morrow. Modigliani, A. 1968. Embarrassment and Embarrassability, Sociometry 3 1:3 13-26.



Moller, H. 1958. The Social Causation of the Courtly Love Complex, Znternatioml Quarterly 1 :137-63. Parsons, T. 1968. On the Concept of Value-Commitments, Sociological Znquiry 38:13 5-60. Rapaport, D. 1942. Emotions and Memory. Baltimore, Md.: Williams and Wilkins. Rubin, Z . 1973. Liking and Loving. New York: Holt, Rinebart & Winston. Sartre, J. P. 1948. The Emotions. New York: Philosophical Library. Schachter, S. 1964. The Interaction of Cognitive and Physiological Determinants o f Emotional States, Psychobiological Approaches to Social Behavior, ed. P. Leiderman and D. Shapiro. Stanford, Cali.: Stanford University Press, pp. 138-73. Scbachter, S., and Wbeeler, L. 1962. Epinephrine, Chlorpromazine, and Amusement, Journal of Abnormal and Social Psychology 65: 121-28. Schoeck, H. 1966. Envy: A Theory of Social Behavior. New York: Harcourt, Brace & World. Seeley, J. R. 1967. The Americanization of the Unconscious. New York: International Science Press. Simmel, G. 1950. The Sociology of Georg Simmel, trans. and ed. K. W O E Glencoe, Ill.: The Free Press. Slater, P. 1964. Social Limitations on Libidinal Withdrawal, The Family, ed. R. L. Coser. New York: St. Martins Press, pp. 220-50. 1966. Microcosm. New York: John Wiley. 1968. The Glory of Hera. Boston: Beacon Press. Walster, E. 1971. Passionate Love, Theories of Attraction and . Murstein. New York: Springer, pp. 85Love, ed. 33. L 99.



Walters, R. H. 1966. Implications of Laboratory Studies of Aggression for the Control and Regulation of Violence, Annals of the American Academy of Political and Social Science 364:60-72. Warner, R. S.; Wellman, D.; and Weitzman, L. 1971. The Hero, the Sambo, and the Operator, paper prepared for the annual meeting of the American Sociological Association, Denver, Colorado. Weber, M. 1966. The Theory of Social and Economic Organization. New York: The Free Press. Wikler, N. 1974. Sexism in the Classroom. Unpublished manuscript, Sociology Dept., University of California at Santa